<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-2506">
  <Title>A Novel Approach to Focus Identification in Question/Answering Systems</Title>
  <Section position="7" start_page="8" end_page="8" type="concl">
    <SectionTitle>
6 Conclusions
</SectionTitle>
    <Paragraph position="0"> Question/Answering and Text Categorization have been, traditionally, applied separately, even if category information should be used to improve the answer searching. In this paper, it has been, firstly, presented a Question Answering system that exploits the category information. The methods that we have designed are based on the matching between the question and the answer categories. Depending on positive or negative matching two strategies allow to affect the Q/A performances: answer re-ranking and answer elimination.</Paragraph>
    <Paragraph position="1"> We have studied five question categorization models based on two traditional TC approaches: Rocchio and Support Vector Machines. Their evaluation confirms the difficulty of automated question categorization as the accuracies are lower than those reachable for document categorization. null The impact of question classification in Q/A has been evaluated using the MRAR and the SRAR scores. When the SRAR, which considers the number of incorrect answers, is used to evaluate the enhanced Q/A system as well as the basic Q/A system, the results show a great improvement.</Paragraph>
  </Section>
class="xml-element"></Paper>